Data Warehouse

This page is aimed at technically experienced users and analytics teams. Here you manage automated data pipelines, precomputed analytical views, and connections to external BI tools (e.g. Metabase, Tableau, Power BI), and define how long certain data is retained.

What can I do here?

  • Create a data pipeline (ETL) that regularly moves data from a source into a destination table
  • Manually trigger a pipeline run and view its execution history
  • Create a precomputed view (materialized view) and refresh it manually if needed
  • Connect to an external BI tool, test the connection, and generate an API key
  • Define retention rules that automatically delete, archive, or anonymize old data
  • View the overall health (active/errored pipelines, storage used) on the overview page

Step by step

  1. Create a pipeline: Switch to the "ETL" tab, click "New pipeline", choose a source type (database, event stream, API, file upload), a destination table and a schedule (cron expression), and save.
  2. Run a pipeline manually: Click the play icon in the pipeline table to trigger an immediate run.
  3. Create a view: Switch to the "Views & BI" tab, click "New view", and enter the underlying query along with a refresh schedule.
  4. Set up a BI connection: In the same tab, click "New connection", choose the BI tool and the allowed tables, test the connection, and generate an API key if needed.
  5. Create a retention rule: Switch to the "Retention" tab, click "New rule", and enter the target table, retention period in days, action (Delete/Archive/Anonymize), and an optional filter condition.

Fields explained

Field Meaning Notes/Impact
Source type (Pipeline) Origin of the data Database, Event stream, API, or File upload
Destination table Table in the warehouse the data is written to Should have a clear, unique name
Schedule (Cron) Repeat interval for the pipeline or view refresh In cron format, e.g. 0 * * * * for hourly
Source query (View) SQL query that populates the view Intended for technically experienced users only
Auto-refresh Determines whether a view refreshes automatically on schedule Can alternatively be triggered manually
BI tool type Connection to Metabase, Tableau, Grafana, Power BI, or Looker Determines the connection profile
Allowed tables Tables the BI tool is permitted to access Limits data access for security reasons
Target table (Retention) Table the retention rule applies to —
Retention days How long records are kept before the action applies —
Action (Retention) Delete, Archive, or Anonymize Determines what happens to expired data
Filter condition Extra condition limiting which records are affected Optional

Values & statuses

Pipeline status

Value Plain meaning
Active Pipeline runs according to its schedule
Paused Pipeline is temporarily halted
Error The last run failed
Draft Pipeline is created but not yet activated

Run status (ETL run)

Value Plain meaning
Running The run is currently in progress
Completed The run finished successfully
Failed The run aborted with an error
Cancelled The run was manually stopped

View status

Value Plain meaning
Active The view is up to date
Refreshing The view is currently being recomputed
Stale The view hasn't been refreshed for a while
Error The last refresh failed

Retention action

Value Plain meaning
Delete Expired records are permanently removed
Archive Expired records are moved to an archive
Anonymize Personal data in expired records is removed, but the record remains

Frequently asked questions

Who is this page for?
For technically experienced users and analytics/BI teams. For everyday use, the "Reports", "Dashboards" and "Data Explorer" pages are usually sufficient.

What happens if a pipeline fails?
Its status changes to "Error" and you'll see an error message in the execution history. You can then manually rerun the pipeline.

How do I protect sensitive data when connecting a BI tool?
Restrict the "Allowed tables" of that connection to exactly the tables the external tool actually needs.